"Violence against women: a statistical overview, challenges and gaps in data collection and methodology and approaches for overcoming them"

Expert Group Meeting

Organized by: UN Division for the Advancement of Women

in collaboration with: Economic Commission for Europe (ECE) and World Health Organization

(WHO) 11 � 14 April, 2005 Geneva Switzerland

Report of the expert group meeting

pdfMachine by Broadgun Software - a great PDF writer! - a great PDF creator! - http://www.pdfmachine.com http://www.broadgun.com

Table of contents I. Background

1.1. Data collection on violence against women 1.2. Objectives of the expert group meeting 1.3. Participants and format of the meeting

II. Summary of the discussion

2.1. Definition of violence against women 2.2. Population-based surveys 2.3. Characteristics of population-based surveys carried out to date 2.4. Gaps and challenges in population-based research on violence against women 2.5. Service-based data on violence against women 2.6. Gaps and challenges with service-based data on violence against women

III. Recommendations adopted by the expert group meeting

3.1 General principles for collecting data on violence against women 3.2 Recommendations for improving population-based surveys 3.3 Recommendations for improving service-based data

***

Discussion on indicators

***

Annex I List of participants Annex II List of documents Annex III Programme of work

1

I. Background On 22 December 2003, the General Assembly of the United Nations adopted by consensus a resolution entitled �In-depth study on all forms of violence against women� (A/RES/58/185). The resolution requests the Secretary-General to conduct an in-depth study on all forms and manifestations of violence against women (the full text of the resolution is attached in the annex). The resolution spells out five areas to be addressed in the study: (i) A statistical overview on all forms of violence against women, in order to evaluate

better the scale of such violence, while identifying gaps in data collection and formulating proposals for assessing the extent of the problem;

(ii) The causes of violence against women, including its root causes and other contributing factors;

(iii) The medium and long-term consequences of violence against women; (iv) The health, social and economic costs of violence against women; (v) The identification of best practice examples in areas including legislation, policies,

programmes and effective remedies, and the efficiency of such mechanisms to the end of combating and eliminating violence against women.

The in-depth study will build on work that has been undertaken so far, synthesize and evaluate findings, and identify good practices and effective strategies, along with gaps and challenges. The study is intended to give a global picture of all forms of violence against women, the scale and prevalence of different forms of violence against women, its causes and consequences, as well as the costs of such violence. The study will identify gaps in knowledge and data collection, and give particular attention to good practice examples to highlight successful strategies to combat violence against women. Importantly, the study will include action-oriented recommendations that encompass effective remedies and prevention and rehabilitation measures. It will also aim to make recommendations for strengthening collaboration between producers and users of data and statistics on violence against women. General Recommendation 19 (1992) on violence against women of the Committee on the Elimination of Discrimination against Women noted that gender-based violence is discrimination within the meaning of article 1 of the Convention on the Elimination of All Forms of Discrimination against Women. The United Nations Declaration on the Elimination of All Forms of Violence against Women (1993) identified three main types of violence against women, namely violence in the family, the general community, and perpetrated or condoned by the State, clarifying that such violence can take physical, sexual and psychological forms. Policy documents such as the Beijing Platform for Action adopted at the Fourth World Conference on Women (1995), the outcome document of the twenty-third special session of the General Assembly of June 2000, entitled �Women 2000: gender equality, development and peace for the twenty-first century�, and various other United Nations resolutions and outcomes elaborated further the forms of violence against women and the sites where it takes place, and actions to combat and prevent its occurrence. These documents also highlighted the ways in which violence against women intersects with, and impacts on, other aspects of women�s well-being and their enjoyment of their human rights.

2

1.1 Data collection on violence against women As increased attention has been paid to the issue of gender based violence, substantial work has been done by governments, United Nations entities, other international and regional organizations, non-governmental organizations and researchers to strengthen survey methodologies, develop common indicators and address the challenges of comparing data over time, and across different countries and regions. However, the lack of data on the nature, prevalence and incidence of the various types of violence against women is regularly highlighted as a major concern and barrier to effective policy making. In March 2004, participants at a high-level round table of the Commission on the Status of Women (CSW) noted that: �In certain areas, such as violence against women, methodological shortcomings and lack of reporting, or under-reporting, led to inaccurate data collection, and such unreliable or misleading information could lead to poor policies. Outdated data sometimes limit the usefulness of the information as a basis for policy-making. At times, there is a discrepancy between the types of statistics produced and the needs of policy makers.�

The Committee on the Elimination of Discrimination against Women, the treaty body established to monitor implementation of the United Nations Convention on the Elimination of All Forms of Discrimination against Women, regularly addresses the issue of data collection, urging reporting States in its concluding comments to strengthen and systematize efforts to gather statistics on the various forms of violence against women, including for example on domestic violence, on trafficked, migrant and refugee women, on sexual exploitation, forced marriages and harmful traditional practices. Governments likewise recognize the need for adequate data on violence against women. For example, in their responses to the Secretary-General�s request for information on implementation of General Assembly resolution 57/179 and 57/181 related to the elimination of all forms of violence against women and crimes committed against women in the name of honour, Member States stressed the importance of data collection and compilation of statistics for policy-making. Given concerns about the lack of reliable, comprehensive and comparable data on various forms of violence against women, and the importance of the availability of such data to policy makers and professionals working to combat such violence and/or support its victims, the study will assess the current situation in the field of data collection regarding violence against women. This will include provision of an overview and evaluation of the quality of available data across countries, an overview of the status of data collection on the different types of violence against women across countries, an analysis of the methodology(ies) currently used to collect such data, as well as an assessment of the remaining gaps and challenges in improving the availability, reliability, comprehensiveness and comparability, in all parts of the world, of such data collection. 1.2 Objectives of the expert group meeting The development and implementation of an effective, comprehensive and multidisciplinary strategy to combat violence against women requires complete, accurate, and up-to-date information on the prevalence and incidence of such violence and its various manifestations. With this in mind, the United Nations Division for the Advancement of Women of the Department of Economic and Social Affairs, in collaboration with the Economic Commission for Europe and the World Health Organization, convened a group of experts in Geneva from

3

11-14 April 2005. The purpose of the meeting was to assess the data and statistics currently available regarding the extent, nature and consequences of all forms of violence against women, evaluate the quality of the data, and identify remaining gaps in data collection. Participants reviewed initiatives currently being spearheaded by international and regional organizations, national and local governments, non-governmental organizations and other entities, and tested in various countries, including national level surveys and the development of common indicators. The current document lays out the expert group�s recommendations for enhancing ways for assessing the extent of the problem, especially through improved and more systematic data gathering in the following areas: a) Assessment of the types of data currently available and collected, including issues such as:

The types of violence against women on which data are collected The scope of data collection (at national/local level, regional or sub-regional coverage) The frequency with which data are collected (systematic, ad hoc) The entities responsible for, or contributing to, data collection (independent entities,

government entities, and non-governmental organizations) The methodologies used to collect and analyze the data (e.g. surveys, secondary

analysis, analysis of collated crime or health statistics) b) Assessment of the gaps in data collection and challenges facing data collectors, including

issues such as: The challenge of under- and non reporting of violence against women Lack of data collection on certain types of violence against women Lack of data by geographical region Methodological problems

c) Assessment of needs of policy makers and professionals working to prevent violence

against women: Constraints faced by data producers, including researchers, government offices, and

organizations of civil society Needs of users, including policy-makers, service providers and activists

d) Recommendations for improved data collection and proposals for better assessing the

extent of the problem. 1.3 Participants and format of the meeting The group consisted of a broad range of experts from both criminal justice and public health perspectives, and included academics, representatives from national and international agencies, activists and policy analysts (see Annex I for the list of participants). The participants elected the following officers: Chairperson: Holly Johnson Vice-chairperson: Ana Flavia d�Oliveira Rapporteur: Mary Ellsberg Facilitator of Working Group I: Patricia Tjaden Facilitators of Working Group II: Ivy Josiah and Holly Johnson

4

During the first two days of the meeting, the experts presented papers and case studies for discussion in plenary. During the third and fourth days the themes presented in plenary were further developed in working groups, and recommendations were developed in regard to each of the topics of the meeting. Following is a summary of the discussions and recommendations developed during the meeting. II. Summary of the discussion The Beijing Platform for Action adopted at the Fourth World Conference on Women categorizes violence against women as an obstacle to the achievement of the objectives of equality, development and peace. It underlines that violence against women violates and impairs or nullifies the enjoyment by women of their human rights and fundamental freedoms. The Platform thus confirms the categorization of violence against women as a human rights issue and the resulting accountability of States to prevent, investigate and punish acts of violence against women, whether perpetrated by the State or by private persons. The consideration of violence against women as a fundamental human rights violation critically enhances earlier approaches, such as criminal justice and public health approaches, to addressing all forms of violence against women. 2.1 Definition of violence against women The Beijing Platform for Action provided an important impetus for data collection and research on violence against women. Specifically, it called attention to the lack of data on violence against women and exhorted governments to build national statistical capacity to collect such data and disseminate findings, and to encourage research into the causes and consequences of different forms of violence against women. It also urged governments to work with universities and other organizations from civil society to develop policies and institutional reforms on violence against women (strategic objectives D2 and H3, paragraphs 129 and 206). Since the emergence of violence against women as an important field of study, researchers and data collectors have been faced with the challenge of developing reliable and valid techniques to define and measure violence against women. Both the definition of violence against women and the means of measuring it vary widely from study to study and from country to country. The 1993 United Nations Declaration on the elimination of violence against women provides a very broad and inclusive framework for studying violence against women. According to this definition,

"Violence against women" means any act of gender-based violence that results in, or is likely to result in, physical, sexual or psychological harm or suffering to women, including threats of such acts, coercion or arbitrary deprivation of liberty, whether occurring in public or in private life.� (General Assembly resolution 48/104 of 20 December 1993, Article 1)

Specifically, the Declaration outlines a broad variety of acts and circumstances that are included in this definition:

5

Physical, sexual and psychological violence occurring in the family, including battering, sexual abuse of female children in the household, dowry-related violence, marital rape, female genital mutilation, and other traditional practices harmful to women, non- spousal violence and violence related to exploitation;

Physical, sexual and psychological violence occurring within the general community, including rape, sexual abuse, sexual harassment and intimidation occurring at work, in educational institutions and elsewhere, and trafficking in women and forced prostitution;

Physical, sexual and psychological violence perpetrated or condoned by the State, wherever it occurs.

The Declaration recognizes that some groups of women are particularly vulnerable to violence, such as: women belonging to minority groups, indigenous women, refugee women, migrant women, women living in rural or remote communities, destitute women, women in institutions or in detention, female children, women with disabilities, elderly women and women in situations of armed conflict. Because it is so inclusive, the Declaration�s definition of violence against women provides a good conceptual framework for studying violence against women. However, because the types of violence vary greatly in their characteristics, the methods for collecting data on violence must be adapted according to the type of violence under study. Particular attention needs to be paid to addressing forms of violence against women which remain underreported and to assessing the prevalence of violence in certain hard to reach populations. The expert group discussed in-depth the strengths and weaknesses of the two most common forms of data collection on the subject � population-based surveys and service-based data. In addition to quantitative data collected through these methods, the group discussed the importance of using qualitative data to study other aspects of violence against women. Qualitative methods can be an especially effective tool for assessing the nature of gender-based violence in an environment where little or no systematic data exists, such as conflict-affected areas. They can also be used to gain deeper insight into the causes and consequences of violence against women and to gauge attitudes towards such violence. Recommendations for how to strengthen the use of these and other methods for collecting data on violence against women are presented in section III of this report. 2.2 Population-based surveys Population-based surveys that query women drawn from representative samples about their experiences as victims of violence are the most reliable method for collecting information on the extent of violence perpetrated against women in a general population. Unlike data that are generated from other methods, such as administration records, in-depth interviews, and observation, population-based surveys obtain information from randomly selected samples. Thus survey results may be generalized to the overall population from which the sample was selected. Because they reflect actual occurrences of victimization rather than what is reported to officials, population-based surveys that query respondents about victimization also more adequately reflect the true levels of violence perpetrated against women. This makes them particularly useful for measuring the extent of violence against women, monitoring trends over time, building awareness, and developing policy.

6

Several governments routinely conduct population-based victimization surveys. For example, the United States of America has been conducting the National Crime Victimization Survey (formerly known as the National Crime Survey) since the 1960s, and the United Kingdom has been conducting the British Crime Survey since 1982. The first population-based survey on violence against women was conducted in the United States in 1975 and focused on intimate partner violence. Since then, numerous independent population-based surveys specifically focusing on one or more types of violence against women have been conducted in various countries, including Australia, Cambodia, Canada, Colombia, Egypt, Finland, Nicaragua, Sweden and Zimbabwe. These surveys all show that the prevalence of violence perpetrated against women is greater than official crime or health data indicate. The second half of the 1990s, particularly in the wake of the adoption of the Platform for Action by the Beijing World Conference on Women in 1995, witnessed a great expansion in the number of surveys conducted to estimate the prevalence of different forms of violence against women, particularly intimate partner violence. In the decade since the Beijing Conference enormous progress has been made in documenting the extent and nature of violence against women. For example, a recent literature review found over 80 prevalence studies from more than 50 countries on intimate partner violence (Ellsberg & Heise, 2005). These studies indicate that between 10 percent and 60 percent of women who have ever been married or partnered in an intimate relationship have experienced at least one incident of physical violence from a current or former intimate partner. Most studies estimate a lifetime prevalence of partner violence between 20 percent and 50 percent. Although women can also be violent, and abuse exists in some same-sex relationships, data from national crime surveys, the United States National Violence Against Women Survey, police, hospital and court records, and clinical and shelter sample surveys show women are overwhelmingly the victims of intimate partner violence and significantly more likely than men to be injured during attacks by intimates. Sexual assault and abuse of women by non-intimate partners has also been proven to be much more common than previously thought. Surveys find that between 20 and 40 percent of women have experienced sexual assault by men other than partners in their adult lifetime. Surveys also show that experiences of sexual violence often begin at an early age. A review of 25 studies worldwide indicates that 11 to 32 percent of women report that they experienced childhood sexual abuse. The collection of data that has been conducted on violence against women has been carried out by a wide variety of agencies, including international agencies, government ministries, statistical offices, universities, non-governmental organizations and women�s rights organizations. These studies have been instrumental in convincing legislators and policy-makers about the high levels of violence against women, the needs of victims, and the urgent requirement for an effective response to such violence. Conducting these kinds of studies is therefore important, particularly in areas where little or no data on the extent of the problem exists. 2.3 Characteristics of population-based surveys carried out to date The surveys that have been carried out to date vary widely with respect to the following:

Types of violence: Population-based surveys have examined many different types of violence against women, including intimate partner violence, sexual abuse and rape,

7

incest, emotional abuse and female genital mutilation/cutting. While most studies focus on just one type of violence (e.g., rape, intimate partner violence, incest), a few examine multiple forms of violence. Intimate partner violence and sexual assault are the most commonly studied forms of violence against women. The term �domestic violence� is often used interchangeably with intimate partner violence, although in some settings, it is also used to refer to other types of violence occurring in the household, such as child abuse, elder abuse, abuse of domestic workers, or violence committed against women by members of the extended family, such as in-laws. It is also important to note that most surveys include a categorization for severity and frequency, often referred to as chronicity, whether directly addressed or embedded in the operationalization of the violence.

Specific types of violence against women have been defined and operationalized in different ways across different surveys. For example, some studies use a legal definition of victimization, while others allow respondents to self-define themselves as victims. Further, some surveys use a single direct question to query respondents about possible victimization, while others use multiple behaviourally specific questions. Multiple behaviourally specific questions, (e.g., �Have you ever been slapped, kicked, or beaten?�) are generally considered more effective for eliciting disclosure on violence than questions that ask women to self-identity as victims (e.g., �were you ever abused?�). Less attention has been given to other forms of violence, such as economic violence, stalking, dowry-related violence, crimes committed in the name of honour, violence in armed conflict settings or trafficking of women for sexual exploitation. (Notable exceptions are the United States Violence against Women Survey, the British Crime Survey, and the Canadian General Social Survey on Victimization, all of which addressed stalking very comprehensively.) Because these forms of violence occur less frequently, or in specific populations they are difficult to study using population-based methods, and are best addressed through other methods.

Prevalence: The majority of studies provide estimates on the prevalence of violence,

that is, the percentage of persons within a demographic group, e.g., women over the age of 18, who are victimized within a specific time frame, such as the person�s lifetime, or in the previous 12 months.

Incidence: Another way to measure violence is to count the number of separate

victimizations or incidents of violence perpetrated against persons within a specific demographic group during a specific time frame. These incidents can be presented as an incidence or victimization rate, which is obtained by dividing the number of victimizations committed against persons in a demographic group by the number of persons in that group and setting the rate to a standard population base, such as 1,000 or 100,000 persons. Some types of violence are more difficult to classify according to incidents, since a women may experience several acts of one of more types of violence at one time.

Reference period of victimization: The more common time frames for measuring

the prevalence and incidence of violence against women are lifetime, since adulthood (e.g., since age 15, 16 or 18), the previous 5 years, previous 12 months, or the previous 6 months.

8

Frequency/duration: Many studies include some measure of frequency (number of incidents), others also examine duration of the violence. Together, frequency and duration are sometimes referred to as chronicity.

Severity: Many studies include measures of severity, determined either by actual

injuries or by classifying the acts according to their risk of injury. Some studies have used other types of measures to classify severity, such as violence during pregnancy, whether the woman feared for her life, the emotional impact of the violence, being unable to work or care for children as a result of the violence, the number of violent incidents, or physical and sexual violence together.

Perpetrators: The relationship between victim and perpetrator is a key category for

classifying the types of violence. The majority of surveys measure intimate partner violence, which is defined as violence perpetrated by current and former spouses, co-habiting partners, dates, and boyfriends. Some studies include violence by same-sex intimate partners. Other surveys, particularly, the International Violence Against Women Survey (IVAWS) and the crime victimization surveys, include violence by a variety of perpetrators, including strangers, family members other than spouses, authority figures (e.g., teachers, police officers, religious leaders, doctors) and acquaintances. The DHS module, while focusing on intimate partner violence, also measures violence by other perpetrators. The same is true for the WHO studies.

Setting: Some surveys specifically look at violence committed in different settings,

such as the home, schools, workplaces, and situations of armed conflict.

Dedicated versus general surveys: Some studies are specifically designed to measure various forms of violence against women (dedicated surveys), while others are surveys designed primarily for other purposes that have included some additional questions on violence against women (general surveys). The number and depth of questions added to general surveys varies considerably. There is also a hybrid form in which a separate module on violence against women is appended to a general survey.

Modes used in administering the questionnaire: In resource- poor settings, the

great majority of surveys are conducted as face-to- face interviews. In more developed settings, postal surveys, telephone interviews, and self-administered questionnaires, often incorporating computerized technologies such as CATI (computer-assisted telephone interviewing), CAPI (computer-assisted personal interviewing), CASI (computer-assisted self interviewing), are commonly used.

Questionnaire design: There is a great deal of variation in the design of

questionnaires used in surveys and this ultimately affects the comparability of the data across different surveys. The wording and the number of questions used to identify women who experienced violence affect the quality of the results and in particular the validity of the estimates of the number of women who have experienced violence.

Study population: Population-based violence against women surveys vary widely

with respect to the age and marital status of the population being examined. Surveys that collect information on health outcomes often interview only women of reproductive age (15 - 49 years), whereas other surveys include all women over a certain age (15 or 18 years). Some surveys, particularly in industrialized countries,

9

interview both men and women about their experiences as victims of violence. Some studies, particularly in the United States of America and Europe, have focused on the experiences of children and youth as victims of violence.

Coverage: Some studies are national in coverage, whereas others are representative of

one or more regions or local communities.

Study organizations: Relatively few studies are carried out under the auspices of National Statistics Offices. More often, they are conducted by universities, independent research institutes and non-governmental organizations (often under the auspices of or with funding from government or international agencies). In some countries where National Statistical Offices are not directly involved in conducting the survey, they have provided assistance with sampling methodology.

While a great deal has been accomplished already, there are still challenges and gaps in developing knowledge on violence against women in all parts of the world. Problems exist with respect to both the reliability and validity of the data being collected, as well as with the comparability of data across studies. There is a lack of standardized methods and instruments for data measurement and collection. Thus, more work is needed to ensure greater uniformity and comparability in the collection and reporting of data on all forms of violence against women. For example, many of the prevalence estimates for intimate partner violence are not comparable, due to methodological differences in the way that violence has been defined and measured, although recent international efforts such as those of the WHO and the IVAWS have sought to address the issue of comparability. Besides the issue of comparability and quality of data, there are enormous gaps in terms of territorial coverage, the populations addressed, and types of violence against women measured. 2.4 Gaps and challenges in population-based research on violence against women Ethical and safety issues The last decade has shown that, despite the sensitivity of the topic, it is possible to collect reliable and valid information on violence against women. However, there is also ample evidence to suggest that specific safeguards are needed to protect the safety of both respondents and field staff. Failure to adhere to these measures can compromise the quality of the data, and also put participants at risk of physical or emotional harm. The World Health Organization has developed safety and ethical guidelines for conducting research on domestic violence1 and on trafficking2 (see box below). These guidelines primarily address issues such as guaranteeing the privacy and confidentiality of the interview, providing special training on gender issues and violence against women to interviewers who will be collecting data, providing a minimal level of information and/or referrals for respondents in situations of risk, and providing emotional as well as technical support for field staff.

1 WHO Putting Women First: Ethical and Safety Guidelines for Research on Violence against Women, Geneva: WHO, 2001. (Also available in Spanish and French) 2 WHO, LSHTM and EC. WHO Ethical and Safety Recommendations for Interviewing Trafficked Women. Geneva:WHO, 2003 (Also available in Armenian, Bosnian, Croatian, Japanese, Rumanian, Norwegian and Serbian),

10

WHO Ethical and Safety

Guidelines for Researching Domestic Violence against

Women

The safety of respondents and the research team is paramount and should infuse all project decisions.

Prevalence studies need to be methodologically sound and to build upon current research experience about how to minimize the under-reporting of abuse.

Protecting confidentiality is essential to ensure both women�s safety and data quality.

All research team members should be carefully selected and receive specialized training and ongoing support.

The study design must include a number of actions aimed at reducing any possible distress caused to the participants by the research.

Fieldworkers should be trained to refer women requesting assistance to available sources of support. Where few resources exist, it may be necessary for the study to create short-term support mechanisms.

Researchers and donors have an ethical obligation to help ensure that their findings are properly interpreted and used to advance policy and intervention development.

Violence questions should be incorporated into surveys designed for other purposes only when ethical and methodological requirements can be met.

Different approaches in violence research and data collection As previously mentioned, in the last decade, two major approaches have emerged with regard to data collection on violence against women. One approach is the dedicated study that is represented by studies that are primarily designed to gather detailed information on the extent of different forms of violence against women. Another approach is to embed questions about violence against women within an ongoing large-scale study that is designed to generate information about a different subject. A hybrid form is to add a special module onto a general survey. Several countries, particularly in Europe and North America, have carried out dedicated national surveys on violence against women, many of which are modelled on the 1993 Canadian Violence against Women Survey. Examples of these are the National Violence against Women Surveys conducted in Australia, France, Finland, Germany, Sweden and the United States of America.

11

There are also on-going efforts of international organizations and institutes to support the implementation of internationally comparative surveys dedicated to violence against women using standard survey methodology. Two important examples of multi-country efforts of dedicated surveys are: (1) the International Violence against Women Surveys (IVAWS), coordinated by the European Institute for Crime Prevention and Control, affiliated with the United Nations (HEUNI), with inputs from the United Nations Office on Drugs and Crime (UNODC), Statistics Canada, and the United Nations Interregional Crime and Justice Research Institute (UNICRI): and (2) the WHO Multi-Country Study on Domestic Violence and Women�s Health. The IVAWS has been administered in eleven countries to date. Its purpose is to collect nationally representative data on a broad array of violent acts perpetrated by men against women, including physical and sexual violence by intimate partners and other men. These studies are conducted within a crime victimization framework and provide information that is particularly useful for interventions in the criminal justice sector. The WHO Multi-Country Study has been conducted in at least ten countries and collects data on women�s experiences of intimate partner violence, sexual assault and child sexual abuse, as well as a broad range of negative health outcomes commonly associated with violence. The WHO studies typically collect data from two sites, a national metropolitan site as well as a provincial site. Increasingly, questions on violence against women are included into large-scale surveys primarily designed for broader, but related purposes, including by using special modules. For example, national crime victimization surveys conducted by various countries such as Canada and the United Kingdom, as well as the Demographic and Health Surveys (DHS), supported by MACRO International, and Reproductive Health Surveys (sponsored by the Centers for Disease Control and Prevention - CDC) conducted in many countries, have included questions on violence against women. The victimization surveys in Canada and the United Kingdom have added modules to questionnaires about general crime victimization in order to address partner violence, sexual assault and stalking. These modules contain detailed questions such as those used in dedicated violence against women surveys. In some cases only one or two general questions are used, such as �Have you even been beaten by anyone since you were 15/were married? By whom?� However, single question approaches are not recommended, as they have been shown to lead to greater under-reporting. The DHS domestic violence module which has now been used more or less consistently by most DHS countries that have or are collecting violence information, is also quite detailed and allows a fairly in-depth study of the prevalence and inter-generational consequences of violence. The DHS has also used a module of questions on female genital cutting to collect information on female genital cutting/mutilation in at least 16 countries. The World Health Organization questionnaire includes all of the questions included in the DHS module and also addresses other aspects such as risk and protective factors as well as information on the strategies and services that woman use. Based on the questionnaire, WHO has also developed a short instrument for measuring the prevalence of violence against women by partners and non-partners that can be integrated into other studies. The dedicated violence against women studies tend to gather more information about different types of violence and perpetrators, as well as information on circumstances and consequences of violence, including injuries, use of services and women�s responses to

12

violence. They also tend to devote more attention to the interaction between interviewers and respondents and to address more comprehensively issues of safety and confidentiality. Experience to date indicates that prevalence estimates produced using these methods tend to be higher than in non-specialized surveys. A disadvantage to these studies is that they tend to be costly, and difficult for countries to repeat on a regular basis. On the other hand, surveys designed primarily for other purposes can also play an important role in documenting the extent of violence against women, particularly when resources are scarce for conducting dedicated surveys. Moreover, the broad variety of other variables collected in these surveys, such as reproductive and child health outcomes, can be used to deepen understanding of risk factors and health consequences, and to monitor violence against women, and its intergenerational consequences, over time. A major disadvantage of embedding violence against women surveys in a general survey designed for other purposes is that the breadth of information generated on violence against women can be more limited than the information generated by dedicated studies. There is also a greater risk of under-reporting, for the reasons mentioned above. In sum, there are trade-offs both in utilizing the dedicated survey and general survey approach, and each can contribute valuable evidence for guiding interventions. For the non dedicated studies, a module/instrument with multiple questions is recommended, as the use of only one or two questions has been demonstrated to result in greater under-reporting. Hard-to-reach populations Significant progress has been made in the development of large-scale surveys to estimate the prevalence and impacts of violence against women in both developed and developing countries. These surveys provide reliable and valid estimates, usually at the national or regional level. New methods and innovations are now needed to address forms of violence that are less common, or occur in specific populations, such as: female genital mutilation/cutting dowry-related violence other traditional or cultural practices that are harmful to women and girls economic violence violence against women in educational or custodial institutions violence against migrant and undocumented workers, refugees, and minority women violence in armed conflict settings trafficking in women for sexual exploitation crimes committed in the name of honour femicide, including murders of women by intimate partners as well as others. Minority or refugee women tend to be omitted from large national surveys that are conducted only in households in the dominant local language. Population-based household surveys may not be the best method for addressing forms of violence against women that are relatively rare, or that occur in very specific populations that may not be captured in overall sampling. Examples of efforts to assess the extent and effects of gender-based violence in refugee, internally displaced and post-conflict settings are summarized in the box below.

13

Reproductive Health Response in Conflict (RHRC) Consortium GBV Initiative From 2000 to 2004 the Reproductive Health Response in Conflict (RHRC) Consortium spearheaded a global �GBV Initiative,� which resulted in, among other outcomes, a qualitative assessment of gender-based violence in conflict settings. The initiative found that there was scant data available about the prevalence of gender-based violence or best practices for qualitatively or quantitatively describing the problem. In an effort to improve international capacity to measure, prevent, and respond to gender-based violence among refugees, internally displaced persons, and other conflict-affected populations, the GBV Initiative designed a �Gender-based Violence Tools Manual for Assessment and Program Design, Monitoring, and Evaluation� (2004). In addition to a series of qualitative assessment tools and techniques, the GBV Tools Manual contains a working draft of a standardized population-based survey designed to measure multiple forms of gender-based violence in conflict-affected settings around the world. Isis-WICCE (Women�s International Cross Cultural Exchange) - Bridging the Gap Realizing that there was very limited information about the realities of women in situations of armed conflict, Isis-WICCE took a deliberate step to utilize two international instruments (the 1993 Vienna Declaration and Programme of Action and the 1995 Beijing Platform for Action) as its framework for documenting violations of women�s human rights in situations of armed conflict in Africa, with Uganda as a starting point. In Uganda, Isis-WICCE collected data on women�s experiences in situations of armed conflict in seven districts. In addition to documenting their experiences, Isis-WICCE also initiated emergency and long-term interventions in support of victims, in the belief that the purpose of data collection is to utilize the information to make a difference in the plight of women. Examples taken from papers presented at the expert group meeting, available at http://www.un.org/womenwatch/daw/egm/vaw-stat-2005/documents.html.

2.5 Service-based data on violence against women Service-based data refers to information that is collected routinely through public and private agencies that may come into contact with abused women, even if the violence is not the primary reason for the contact. They include records from health centers, police stations and courts, public services such as housing or social welfare services, and shelters and other support services for survivors of violence. Examples of other support services include women lawyers� associations, legal aid services and advocacy organizations. The information collected by services cannot be used to measure the prevalence of violence in a community since in most societies very few abused women actually report violence to the police or other services, and those that do report tend to be the most seriously injured and marginalized women. However, these data, where they exist, can make a valuable contribution to understanding the causes and dynamics of violent acts and societal responses to violence. Service-based statistics can be used to monitor the number of women coming forward to various agencies for help, and can identify the scope of violence-affected women among specific populations, such as those presenting to medical centres with injuries, or abused women as a percentage of all assaults presenting to criminal court. Together, these data can

14

contribute to estimates of the cost to society of responding to violence against women. They can also be used to estimate the need for training among service providers, including medical and criminal justice professionals, concerning the effects of violence on women and how best to respond to reduce trauma and increase safety. In addition, service-based statistics, if collected in sufficient detail, can contribute to evaluations of the response of these agencies to abused women. It is important to know, for example, how police respond when a woman comes to them for help, whether the case is investigated, an arrest is made, or charges are dropped. These data are important for monitoring the effectiveness of agencies to whom women turn for help and to increase accountability. Data from police and courts are also needed to evaluate and formulate legislation, procedures and policies to respond to violence. Tracking the availability of services for abused women, for example shelters or refuges and other support services, is also needed to evaluate a society�s response to the problem. These data also provide important context to analyses of the numbers of women coming forward for help. For example, growth in the availability of services may explain growth in the numbers coming for help. By the same token, low numbers of women using shelters or other services should not be interpreted as low demand or need in areas where few such services exist). The following section identifies service agencies with the potential to contribute statistical data about women seeking help or redress for violence. Areas of service delivery included are health, criminal and civil justice services, public services, and refuges and other support services for survivors of violence. Health services: The health setting provides an opportunity for women exposed to

multiple victimizations to be identified in order to receive treatment, care and support and to prevent further violence. However, it is well known that women often do not disclose, even when violence is the underlying cause of their health care visit. One way of increasing disclosure is through routine enquiry about violence, but the effectiveness of this approach in improving outcomes for women is still under discussion. Where routine enquiry takes place, the health service needs to have the capacity to record, analyse and report data on violence against women and its consequences.

There is considerable debate about whether health workers who identify victims of violence should be obliged to report cases to the criminal justice system (�mandatory reporting�). Many health workers believe that this is a breach of privacy and confidentiality and can result in lower disclosure and increased risk for women. In addition, health care workers have also raised ethical concerns about routine surveillance that identifies women who need help when they are not able to provide the appropriate assistance, i.e., where services are not available, or are overwhelmed by need. It is well known that many more cases occur than are reported to either police or health services. However, surveillance (i.e. routine data collection) of specific health outcomes related to violence, such as injury, death from homicide and other data from medico-legal clinics where forensic examinations take place, all offer potential for monitoring trends in violence against women, particularly intimate partner violence and sexual assault by partners and other perpetrators.

15

Criminal and civil justice services: Criminal and civil justice statistics may be collected more systematically than statistics in some other sectors because these agencies operate on the basis of a code of law. It should therefore be possible to organize data by criminal code sections. This data collection would be much enhanced if it were cross-classified with information to ascertain if these crimes were domestic or against women. Some countries have specific laws on domestic violence while others have laws on assault, grievous bodily harm, sexual assault, stalking, homicide and other crimes. Justice sectors have the potential added advantage of being able to collect information on both victims and perpetrators, and to track repeat victimization and repeat offending. Police may also be the primary source of information on intimate partner homicides and other types of femicide.

Criminal courts are also highly structured and operate based on a country�s codes of law. Although cases brought in criminal court represent a very small and non-representative sample of cases of violence against women, court statistics can contribute to an understanding of the response of courts to cases of violence brought before them, and in particular the effectiveness of laws designed to protect women. The civil justice sector is also used by women escaping domestic violence. In some countries civil injunctions, also known as peace bonds, restraining orders, or domestic violence orders, order violent partners away from coming into contact with the victim and can include other conditions, such as drug/alcohol prohibitions or weapons prohibitions. Other types of injunctions can have the violent partner removed from the home. While criminal justice agencies are available as a support to victims of violence, it is also a reality that police can also be responsible for committing violence against women who come to them for help.

Other public services: Most providers of services routinely keep some statistics on the

use of their services, but these services vary in the quality and quantity of data collection, both in general and in relation to violence against women. Such public services may include agencies providing public housing, child welfare or other types of social services.

Public housing The best data collection in housing agencies occurs where there is a statutory/legal duty on public housing providers to prioritize the housing needs of those made homeless due to domestic violence, since they need to know whether their clients are survivors of domestic violence or not. In some countries this leads to the provision of data at a national level of the numbers of those in public housing who have been subject to domestic violence, in an effort that can be considered as an example of good practice. Social services In some countries social workers record in case files whether or not domestic violence and other forms of violence against women has occurred in the family or household that they are supporting for other reasons (e.g. supporting families where there is a risk of child abuse). In some cases it may be appropriate to recommend that this data collection is made more systematic and that aggregate statistics are made available at a national level, but this is dependent upon being able to ensure appropriate safeguards and confidentiality for the women involved.

16

Refuges and other support services Shelters/refuges; rape/sexual assault phone lines; advocacy and related support services; research and documentation centres A variety of support services led by non-governmental organizations, sometimes with some support from public funds, collect information on the extent and nature of violence against women. These include shelters and refuges, sexual assault phone lines, advocacy and related support services. In addition, information is sometimes collected by women�s lawyers associations and legal aid services. Research and documentation centers also collect quantitative and qualitative information on various forms of violence against women. Records of some sort are kept, but these vary considerably in range of information and its quality. The information collected in these services is particularly relevant for qualitative research.

The National Centre for Human Rights (NCHR) - Jordan The National Centre for Human Rights, established in 2003, is entrusted with monitoring the status of all human rights in Jordan and documenting violations. In order to monitor the status of women in Jordan, it is crucial that accurate and regularly updated data is collected, classified, and analyzed with utmost transparency. This is an important requirement for examining the status of women in general and for understanding issues of violence against women in particular.

Several, albeit limited sources of quantitative data are available, including mortality rates � the most widely collected and readily available data on violence against women worldwide. All other data on violence against women is not collected systematically. Medical reports used for a study conducted by the National Centre for Forensic Medicine indicates that during the years 1990-1993, �domestic violence� represented 3.4% of the total reported cases of violence in Jordan. Of these cases 97% were women who were physically abused by their husbands. A 1998 study of official figures from the mid-1990s showed that, at that time, murder was the most frequent crime against women, and that �honor crimes� (including murder, attempted murder, and �accidental murder�) accounted for the largest category of all homicides of women.

A number of sources collect different types of quantitative data. The Women's Safe House developed and managed by the Jordanian Women's Union (JWU) records the number of women who use the shelter's services, as well as its legal and social counseling services. The Family Protection Department collects quantitative data regarding cases reported to the main branch in Amman Governorate and is compiling data from the newly established branches in the other governorates.

17

Bangladesh National Women Lawyers Association (BNWLA)

A study on violence against women was recently carried out in Bangladesh as part of the WHO Multi-Country Study on Women�s Health and Domestic Violence. In addition, there are two primary sources of data on gender-based violence: court records of cases (prosecuted) and police records on reported crimes. Police headquarters established a cell on violence against women complaints in 1993. Data on complaints made by women are collected from all the police stations and from the district women�s affairs cells. Hospital records and newspaper records are also regarded as important sources for collection of data. Statistics on gender-based violence are also collected by the six divisional offices of the Ministry of Women�s and Children�s Affairs. Aside from official statistics, several non-governmental organizations accumulate data by conducting research. This research covers selected areas rather than the whole country. Several non-governmental organizations compile data based on newspaper records. Newspapers have a tendency to cover sexual crimes rather than domestic violence cases. As a result, reports of maternal mortality or suicide get less attention and adultery, child marriage, forced marriage and forced prostitution receive little coverage unless there is an alleged murder.

In order to assist with the accurate and systematic collection of data on violence against women, the Bangladesh National Women Lawyers Association recommends that computerized databases be maintained by the police department, local governments, non-governmental organizations, the judiciary and the One Stop Crisis Center. The various databases could be integrated into a centralized national database. The BNWLA also recommends that community based studies on the prevalence of violence against women be routinely carried out and that data on violence against women be made available to organizations and individuals who wish to access this information. Examples taken from papers presented at the expert group meeting, available at http://www.un.org/womenwatch/daw/egm/vaw-stat-2005/documents.html.

2.6 Gaps and challenges with service-based data on violence against women The availability and quality of service-based data varies a great deal among settings. In some countries, particularly in Europe and North America, information is available from a broad array of sources, although even in these regions additional efforts are needed to collect and present violence-related statistics. In other areas, particularly in resource-poor settings, the challenges for collecting service-based data on violence are much greater. In general, service agencies do not have data collection as their primary responsibility and data available through these agencies are often not collected in a systematic way and consequently exist in a non-standardized form. The quality of the data may be weak, inconsistent over time and not entirely representative of the populations they are attempting to describe. Double counting is a common problem whereby women seeking services from the same agency or from more than one agency over time are counted more than once, resulting in inaccuracies in estimating the total number of help-seeking women. These problems result from inadequate training, resources and capacity for those collating the data, poor coordination among agencies and poor support.

18

Problems with service-based statistics are compounded when social services apart from the police do not exist, exist only in limited number, or when violence against women is not recognized as a crime by police or by society in general and is rarely acted upon. In addition, in certain societies, victims of domestic violence or sexual assault are highly stigmatized and victims therefore rarely come forward for support. As a result, considerable variation exists in the extent to which service-based statistics are available at the country level. III. Lessons learned and recommendations for improving the quality and availability

of data collection on violence against women 3.1 General principles for collecting data on violence against women Substantial progress has been made in the past ten years in the development of methodologies and procedures for data collection on violence against women and a reliable body of evidence on the extent, nature and consequences of violence is currently available. However, this data is difficult to compare across countries and regions. Likewise, tools for active data collection on violence against women are now widely available and need to be applied and used consistently to strengthen the knowledge base for action against violence against women. Data collection, together with qualitative research and analysis on violence against women, can be a form of prevention and intervention. Such research and data collection must be geared towards action: insights gained must form the basis for policy development and implementation, as well as the provision of support and services to victims of violence. The expert group meeting recommends that the following general principles should guide the data collection process:

National statistical agencies, or relevant ministries such as health or justice, have an

important role to play in setting standards, ensuring consistency of concepts, ensuring regular frequency of data collection, and ensuring that data are widely disseminated in a timely fashion. National machineries for the advancement of women which serve as the central policy-coordinating unit on gender equality inside government (see Platform for Action, section H3), should be closely associated with such efforts.

Data collection activities and methodology need to be developed in consultation with a

wide range of stakeholders, including data suppliers, advocates and agencies providing services to women, policy-makers, legislators and researchers to ensure that the data are relevant to stakeholders. This will necessitate multi-sectoral, interagency coordination of the development, implementation, monitoring and evaluation of data collection initiatives.

Data collection and dissemination needs to be transparent. Data should be disseminated in

aggregate format as broadly as possible. Women�s safety is a priority and must not be jeopardized at any point during data

collection. Confidentiality and privacy is of utmost importance both during data collection and in the

processing and storage of data. Safe storage of data to protect the respondents must be a priority.

19

The expert group recommends that comprehensive data collection on all forms of violence against women should be performed in accordance with the mandate of the Beijing Platform for Action, the Fundamental Principles of Official Statistics as agreed by all Member States and the WHO Ethical and Safety Guidelines for Researching Violence against Women.

Specifically the expert group meeting recommends: There is the need to strengthen the capacity of countries to systematically and accurately

collect, process and disseminate data on violence against women, and to use such data as a basis for effective and targeted legislative, policy and programme development and implementation on all forms of violence against women.

All relevant stakeholders including women's organizations, non-governmental

organizations, other service providers and policy makers should be included in all stages of design and process of data collection and use. Strengthening collaboration between producers and users by having the users on board from the beginning of development and implementation of data collection processes is essential for creating ownership and use of the data for interventions.

A variety of research methods and data collection procedures can be used depending on

resources, need for information, and data gaps. Statistical data should be complemented by qualitative methods to capture the complexity, variety and depth of women's experiences with violence.

3.2 Recommendations for improving population-based surveys International experience has shown that the definition and methods used to measure violence can significantly affect prevalence and incidence estimates. In order to improve the quality of data collection on violence against women, the expert group meeting recommends:

National governments should encourage further research and data collection on all forms

of violence against women to be carried out by a variety of actors, including national statistics offices, government agencies, research centres and universities, non-governmental organizations and international agencies.

Methodological guidelines should be developed, compiled and disseminated to assist

countries in designing, in a sustainable way, their national surveys on violence against women.

Data should be collected, processed and disseminated in accordance with the

Fundamental Principles of Official Statistics. When possible, data should be made available for use by researchers and policy-makers after the research is completed.

Surveys should use multiple methods to measure violence, and should include a variety of

different types of violence, including physical, sexual, emotional and economic violence, as well as violence by different perpetrators, institutional violence and traditional practices that are harmful to women. Multiple behaviourally specific questions should be used when possible to elicit information about violent victimization.

20

Multiple approaches and measures should be used to determine the severity of violence. These include the extent of physical as well as emotional injuries, frequency of violence (incidents); and outcomes such as reproductive health problems, economic and social participation, costs and service utilization, consequences for children and child mortality. Different outcomes will be of special interest for different constituencies (for example, injury statistics are particularly important for the criminal justice sector, whereas the impact of violence on mortality and fertility are of interest to the health sector).

Researchers and national statistics offices should make every effort to minimize risks to

respondents and interviewers, under the basic ethical principle of �Do no harm� When such efforts are not feasible, data collection should not be performed.

Types of violence should be presented in a differentiated way and data collected

accordingly so as to enable more nuanced understanding of the range of violence women experience, the consequences for victims and the implications for policy-makers.

When collection of data on violence against women is included into studies designed

primarily for other purposes, a sufficient number of questions should be included to provide a valid measure of violence, and specific measures should be used to maximize the protection of respondents and interviewers. Ideally, a specific module on violence against women should be used rather than a few questions. If basic conditions are not available to enable the collection of reliable estimates of violence against women that takes into account the safety of women, then it is advisable not to conduct the research.

Efforts should be made to strengthen national statistical and research capacity for

collecting data on violence against women, through both specialized surveys and routine data collection.

Findings should be made available in a variety of ways and formats accessible for various

audiences. Data should not be suppressed or censored. Those actors who collect data on violence against women have an obligation to make the findings available to the government and civil society, to use data for awareness raising and the development of programmes and policies for prevention and treatment of all forms of violence against women.

3.3 Recommendations for improving service-based data In order to improve the quality and usefulness of data that may be available through the justice system, the expert group meeting makes the following recommendations: Police-based data collection systems need to be strengthened in all countries worldwide.

At a minimum, countries should aim to develop a statistical recording system that identifies in each case of violent crime reported to them:

Characteristics of victims and offenders: the relationship between the victim and offender (including for example spouse, ex-spouse, dating partner, other relative, acquaintance and stranger), sex of victim and offender, age of victim and offender, repeat victimizations, repeat offenders.

21

Offence characteristics: legal code violation (for example, assault, grievous bodily harm, homicide, sexual assault, stalking), method of inflicting injury, and level of injury. Detailed data collection on homicides: the relationship between the victim and offender (including for example spouse, ex-spouse, dating partner, other relative, acquaintance and stranger), sex of victim and offender, age of victim and offender, location, method of killing, motive for killing (such as jealousy or dowry-related). Outcome of cases: whether the case is investigated, an arrest is made, charges are dropped or withdrawn, what referrals were made to refuges or other services for victims.

Court-based statistics also need to be strengthened in all countries in order to better

document the court response to cases brought before them. Countries should aim for the following minimum set of data elements to be incorporated into and routinely collected in criminal court data collection systems:

Characteristics of victims and offenders: the relationship between the victim and offender (including for example spouse, ex-spouse, dating partner, other relative, acquaintance and stranger), sex of victim and offender, age of victim and offender, repeat victims and offenders coming before the court. Offence characteristics: legal code violation (for example, assault, grievous bodily harm, homicide, sexual assault, stalking). Outcome of cases: charges are dropped or withdrawn, plea bargaining to a lesser charge, convictions, guilty pleas and acquittals, type of sentence, bail granted or denied.

Improved record-keeping is needed in the area of civil injunctions in order to better

monitor the number of such actions being ordered, in addition to violations of these orders.

Efforts should be strengthened to identify and monitor through data collection the

incidence of violence against women perpetrated by police. In order to improve the quality and usefulness of data that may be available through health services, the expert group meeting recommends: Data collection should not be seen as a primary goal for health services and should not be

carried out in isolation from the provision of quality services to survivors of violence. Therefore, capacity building needs to ensure not only capacity of health information systems, but should encompass efforts to increase awareness and capacity among health workers to identify, treat and refer survivors of violence.

Health systems should use internationally agreed standards (International Classification

of Diseases and Related Health Problems, Tenth Revision or ICD-10, and the International Classification of the External Causes of Injury (ICECI), and the International Classification of Functioning Disability and Health (ICF) for classifying and coding diseases and health consequences, as well as deaths.

22

To build capacity for hospitals and clinics to be able to implement injury surveillance systems and ensure that the relationship of victim to perpetrator is included as one of the mandatory variables.

Health services that routinely enquire about violence against women (e.g. antenatal care,

mental health, etc) should record this data in a complete and accurate manner, while ensuring confidentiality. The reporting on violence against women should be incorporated into ongoing service utilization and surveillance systems, rather than implemented as a stand-alone reporting process.

In order to improve the quality and usefulness of data that may be available through other public services, the expert group meeting recommends: All other public services that come into contact with women victims of violence, such as

public housing or social services, should strengthen efforts to collect statistical data in a systematic way and disseminate these data publicly. This includes data on the number and characteristics of women using each service as well as the number of services available to them.

* * *

Discussion on indicators The expert group meeting held a brief discussion on the question of indicators on violence against women. Due to time constraints, the group did not attempt to prepare recommendations. Instead, it agreed to reflect a summary of discussions in the report of the meeting, highlighting the main issues raised. Experts noted that indicators had played an important role in the development of gender equality policies in various fields. Indicators on violence against women enhance the capacity to evaluate policies by making it possible to provide hard data on whether or not violence has increased or decreased as a result of their implementation. Even where data were not available from the outset for a new indicator, agreement on its use also increased the pressure to collect and produce the relevant data. Experts noted that the opportunity afforded by the expectation to measure performance towards achievement of the Millennium Development Goals (MDGs) should be fully utilized to enhance measurement in the area of gender equality. The Millennium Declaration enjoyed the highest political support and commitment, and its review process in the fall of 2005 should be used to push for inclusion of one, or several indicators on violence against women in the MDG indicator set. Agreement on an indicator would be especially useful as considerable efforts had been made in many areas to develop and enhance statistical services and data collection since the issuance of the MDG indicator set in 2000 at a time when data were lacking in many countries and areas to comprehensively measure and track progress in the 48 existing indicators.

In regard to violence against women, different views were expressed as to what an indicator should actually measure. Some experts proposed to measure changes in prevalence and incidence of violence against women over time, while others suggested measuring the use of services. Advantages and drawbacks of these types of indicators were noted, while caution in regard to the breath of some of the suggested indicators was also suggested.

23

The following suggestions of possible indicators were made:

implementation of population based surveys on a regular basis to track the prevalence and/ or incidence of different forms of violence against women

incidents of violence against women measured through service-based data; if services are available this would give an indication of the impact of violence against women and the need for services

the impact of violence against women training of professionals, such as the police and law enforcement, health care

workers and social workers, on violence against women percentage and extent of punishment of perpetrators budgetary allocations to combat violence against women legislation to combat violence against women, national action plans, extent of

availability of services and other measures improvements in women�s property rights and women�s status under family law amendments of discriminatory laws ratification of the Convention on the Elimination of All Forms of Discrimination

against Women and its Optional Protocol and withdrawal of reservations changing attitudes among the general population and among public servants

towards violence against women.

Experts noted that in general, there was significant need to enhance the disaggregation of existing data on the basis of sex in all data collection, and to increase gender sensitivity especially of statisticians. They also emphasized the critical importance of discussing the development of possible indicators, while also suggesting that in some countries, information may not be available to decide which indicator would best serve the purpose of tracking efforts to combat violence against women. Experts also cautioned that a full picture of violence against women could not be established using just one indicator, arguing in favour of using multiple indicators, for different purposes. However, all agreed that the collection of more data on violence against women was important.

* * *

Annex I List of participants Ms. Elizabeth ARDAYFIO-SHANDORF Director of Family and Development Programme Dept. of Geography and Resource Development University of Ghana P. O. Box 59 Legon, Ghana Phone: 233-21-500786 Email: ardayfel@ug.edu.gh Ms. Asmita BASU Legal Consultant and Project Coordinator Lawyers Collective (Women's Rights Initiative) 63/1, Masjid Road Jangpura - Bhogal New Delhi 110014, India Phone: 91-11-2431-3905 Email: wri.delhi@lawyerscollective.org Mary ELLSBERG, Ph.D Senior Advisor Gender, Violence and Human Rights Program for Appropriate Technology in Health (PATH) 1800 K Street NW, Suite 800 Washington DC 20006 Phone: 202-822-0033 Fax: 202-457-1466 Email: mellsberg@path-dc.org Ms. Sharmeen A. FAROUK Senior Research Officer Bangladesh National Women Lawyer�s Association House # 60/A Road # 27 (Old) Dhanmondi R/A, Dhaka � 1209, Bangladesh Phone: 880-2-812-3060 Fax: 880-2-812-5866 Email: safarouk@yahoo.com Email: bnwla@bdonline.com Ms. Dalia FAROUKI Head, Research and Documentation Unit National Centre for Human Rights P.O. Box 5503 Amman 11183, Jordan Phone: 962-6-593-1256 Fax: 962-6-593-0072 Email: dfarouki@yahoo.com Email: dalia.f@nchr.org.jo

Ms. Dominique FOUGEYROLLAS-SCHWEBEL Chargée de recherché en sociologie CNRS-IRIS-CREDEP-UMR 7170 Université Paris-Dauphine F-75775 Paris Cedex 16, France Phone: 33-1-44-05-4863 Fax: 33-1-44-05-4648 Email: Dominique.Fougeyrollas@dauphine.fr Ms. Holly JOHNSON, Ph.D Chief of Research Statistics Canada (On secondment to:) Australian Institute of Criminology GPO Box 2944 Canberra ACT 2601, Australia Phone: 61-2-6260-9245 Fax: 61-2-6260-9201 Email: holly.johnson@aic.gov.au Ms. Ivy JOSIAH Executive Director Women�s Aid Organization No 15, Jalan 14/28, Section 14, Petaling Jaya 46100 Selangor, Malaysia Phone: 60-3-7957-5636 Fax: 60-3-7956-3237 Email: ivyj@pc.jaring.my Email: wao@po.jaring.my Ms. Sunita KISHOR, Ph.D Senior Gender Specialist Measure DHS, Macro International Inc. 11785 Beltsville Drive, Suite 300 Calverton, MD 20705 Phone: 301-572-0384 Fax: 301-572-0999 Email: Sunita.kishor@orcmacro.com Mr. Sami NEVALA Project Manager of the International Violence Against Women Survey (IVAWS) HEUNI, European Institute for Crime Prevention and Control affiliated with the United Nations P.O. Box 444 Pitkänsillanranta 3 A 00531 Helsinki, Finland Phone: 358-10-366-5288 Fax: 358-10-366-5290 Email: sami.nevala@om.fi Email: sami.nevala@iki.fi

Ms. Ruth Ojiambo OCHIENG Director Isis-Women's International Cross Cultural Exchange (Isis-WICCE) 32 Bukoto Street, Kamwokya P.O. Box 4934 Kampala Uganda Phone: 256-41-543-953 Fax: 256-41-543-954 Email: oruth@isis.or.ug Email: isis@starcom.co.ug Ms. Ana Flávia D�OLIVEIRA, M.D Faculdade de Medicina University of São Paulo Av. Dr. Arnaldo, 455, 2° andar, sala 2245, Cerqueira César São Paulo � SP 01246-903 Brazil Phone: 55-11-3066-7085 Fax: 55-11-3062-6018 Email: aflolive@usp.br Ms. Patricia TJADEN, Ph.D Director, Tjaden Research Corporation 768 Franklin Street Denver CO 80218 USA Phone: 303-333-5403 Fax: 303-331-8021 Email: tjadenp@comcast.net Ms. Sylvia WALBY, Ph.D Professor of Sociology School of Sociology and Social Policy University of Leeds Leeds, LS2 9JT United Kingdom Phone: 44-113-343-4437 Email: S.Walby@leeds.ac.uk Ms. Jeanne WARD Gender-based Violence Consultant, IRIN P.O. Box 1360 Sarit Center Nairobi 0606 Kenya Phone: 254-7-3491-6145 Email: Jeanne@theirc.org Email: Ward_Jeanne@hotmail.com

Organizers UN DESA - Division for the Advancement of Women Ms. Christine BRAUTIGAM Chief Women�s Rights Section Division for the Advancement of Women UN DC2-1228 New York, NY 10017 USA Phone: 1-212-963-0535 Fax: 1-212-963-3463 Email: brautigamc@un.org Ms. Monique WIDYONO Project Officer Division for the Advancement of Women UN DC2-1245D New York, NY 10017 USA Phone: 1-917-367-9030 Fax: 1-212-963-3463 Email: widyono@un.org Economic Commission for Europe Ms. Angela ME, Ph.D Chief Social and Demographic Statistics Section Statistical Division, UNECE Room C.440, Palais des Nations 8-14 Avenue de la Paix 1211 Geneva 10 Switzerland Phone: 41-22-917-1666 Fax: 41-22-917-0040 Email: Angela.Me@unece.org Mr. Enrico BISOGNO, Ph.D Social and Demographic Statistics Section Statistical Division, UNECE Room C.410, Palais des Nations 8-14 Avenue de la Paix 1211 Geneva 10 Switzerland Phone: 41-22-917-1662 Fax: 41-22-917-0040 Email: Enrico.Bisogno@unece.org

Ms. Mijidgombo OYUNJARGAL Statistical Division, UNECE Room C.450, Palais des Nations 8-14 Avenue de la Paix 1211 Geneva 10, Switzerland Phone: 41-22-917-4147 Fax: 41-22-917-0040 Email: Mijidgombo.Oyunjargal@unece.org Mr. Paolo VALENTE, Ph.D Statistical Division, UNECE C.410, Palais des Nations 8-14 Avenue de la Paix 1211 Geneva 10, Switzerland Phone: 41-22-917-3306 Fax: 41-22-917-0040 /0546 E-mail: paolo.valente@unece.org World Health Organization Dr. Claudia GARCIA-MORENO, M.D. Coordinator, Gender, Violence and HIV/AIDS Department of Gender, Women and Health World Health Organization 20 Ave Appia Geneva 27-CH 1211, Switzerland Phone: 41-22-791-4353 Fax: 41-22-791-1585 Email: garciamorenoc@who.int Ms. Henrica A.F.M. JANSEN Epidemiologist Department of Gender, Women and Health World Health Organization 20 Ave Appia Geneva 27-CH 1211, Switzerland Phone: 41-22-791-3106 Fax: 41-22-791-15 85 E-mail: jansenh@who.int Other participants Ms. Claudia CAPPA Assistant Programme Officer Strategic Information, Division of Policy and Planning UNICEF, 3 UN Plaza New York, NY 10017, USA Phone: 1 212 326 7753 Fax: 1 212 735 4411 E-mail: ccappa@unicef.org

Ms. Amaya GILLESPIE, Ph.D Director Secretariat of the UN Secretary General�s study on violence against children PO Box 48 1211 Geneva 20 CIC Switzerland Phone: 41 22 909 5633 Office: 41 22 909 5633 Email: agillespie@unicef.org Ms. Erlinda GO Statistics Division Department of Economic and Social Affairs UN DC2-1550 New York NY 10017 USA Phone: 1-212-963-4507 Email: go@un.org Ms. Riet GROENEN Adviser on Socio-Cultural Aspects and Gender Analysis UNFPA/Country Technical Services Team for Europe and Central Asia Grosslingova 35 811 09 Bratislava Slovak Republic Phone: 421-2-593 37 308 Fax: 421-2-593 37 300 Email: groenen@unfpa.org Ms. Lucinda O'HANLON Human Rights Officer Assistant to the Special Rapporteur on violence against women, its causes and consequences UN Office of the High Commissioner for Human Rights Palais Wilson Geneva Switzerland Phone: 41-22-917-9615 Email: lohanlon@ohchr.org Ms. Linda SABBADINI ISTAT National Institute of Statistics Via Cesare Balbo 16 00184 Rome Italy Phone: 39-06-4673-4606 Fax: 39-06-543-0987 Email: sabbadin@istat.it

Annex II List of documents Papers by experts

Violence against women: The Ghanaian case Elizabeth Ardayfio-Schandorf

Violence against women: a statistical overview, and challenges and gaps in data

collection and methodology and approaches for overcoming them. Needs of users of data collection Indira Jaising (presented by Asmita Basu)

Data on violence against women: The Jordanian situation

Dalia El Farouki

Violence against women : A statistical overview, challenges and gaps in data collection and methodology and approaches for overcoming them Sharmeen A. Farouk

Violence against women in Brazil: overview, gaps and challenges

Ana Flávia d�Oliveira & Lilia Blima Schraiber (presented by Ana Flávia d�Oliveira)

Violence against women in France: The context, findings and impact of the Enveff

survey Dominique Fougeyrollas-Schwebel

Assessing the prevalence of violence against women in Canada

Holly Johnson

Domestic violence measurement in the demographic and health surveys: The history and the challenges Sunita Kishor

International Violence Against Women Survey (IVAWS)

Sami Nevala

The efforts of non-governmental organizations in assessing and documenting the violations of women�s human rights in situations of armed conflict: The Isis �WICCE experience Ruth Ojiambo Ochieng

Defining and measuring violence against women: Background, issues, and

recommendations Patricia Tjaden

Improving the statistics on violence against women

Sylvia Walby

Conducting population-based research on gender-based violence in conflict-affected settings: An overview of a multi-country research project Jeanne Ward

Background papers and briefs: Multiple Indicator Cluster Surveys (MICS)

UNICEF

Linkages between violence against women and girls and UNICEF�s Medium Term Strategic Plan, 2002-2005 Priorities UNICEF

Issue brief prepared by the Center for Women�s Global Leadership �Mujer y Desarrollo: Violence against women in couples: Latin America and the

Caribbean. A proposal for measuring its incidence and trends� 2004 Economic Commission for Latin America and the Caribbean

�Making violence against women count: Facts and figures� 2004

Amnesty International

Excerpts from the responses to the questionnaire on the Review and Appraisal of the Beijing Platform for Action and the outcome document of the twenty-third special session of the General Assembly

�Beijing +10: Progress made within the European Union� Excerpt from the report of

the Luxembourg Presidency of the Council of the European Union for the ten year review of the Beijing Platform for Action

Issue brief prepared by the International Gay and Lesbian Human Rights Commission

Annex III Programme of Work

Monday 11 April 2005 8:30 AM Pick up UN Pass at Security Office (See information note) Registration at Palais des Nations Salle XII 10:00 � 10:45 AM Welcome and introductions by UN Division for the Advancement

of Women, Economic Commission for Europe and World Health Organization

Plenary Session 1 Objective: To review existing initiatives to collect data and statistics on violence against women, covering 3 broad areas: international measurement survey initiatives, nationally initiated surveys, and routine data collection. Key questions: On what forms of violence against women are data collected? What does this data indicate about the prevalence of certain forms of violence against women? What methodologies are used and who collects the data? What are the strengths and weaknesses of the different approaches? What challenges remain in getting accurate and complete data on the prevalence of such violence? On what forms of violence against women are data and statistics not collected or available? In which areas and regions are statistics available and in which are they not? Note to presenters: It is anticipated that all papers to be presented in plenary session will be circulated to participants approximately one week before the meeting. Please note that presentations in plenary should be strictly limited to fifteen (15) minutes and should focus on the constraints and remaining gaps and challenges in data and methodology. The papers may be used to present findings of surveys or reports. 10:45 � 12:30 PM Morning session

Panel 1: Data collection on violence against women

Assessing the strengths and weaknesses of different sources of the data on violence against women collected and available through existing systems such as health, crime, death, social service, shelter and other statistics as well as surveys. What methodologies are used for collecting data? How reliable is the data? What are the challenges in setting up a surveillance system for violence against women? What kind of data could national statistical offices collect and analyze to serve as indicators for violence against women? What are the lessons learned from countries trying different approaches? What are the remaining gaps and challenges. Patricia Tjaden Sylvia Walby

(Presentations will be followed by brief questions for clarification)

Panel 2: International efforts at collecting data on various forms of violence against women through surveys

What methodologies are used for collecting, documenting and analyzing data on the prevalence of violence against women? How good are the methodologies and how reliable is the data that is collected? What do these initiatives indicate about the prevalence of certain types of violence against women? On what forms of violence against women are data collected? What are the strengths and limitations of the different methodologies used? What are some of the gaps that remain in data, methodology and regional coverage? WHO (Claudia Garcia Moreno) IVAWS-HEUNI (Sami Nevala) Macro/ DHS (Sunita Kishor) MICS (Claudia Cappa - UNICEF)

(Presentations will be followed by brief questions for clarification)

12:30 � 2:00 PM Lunch 2:00 � 6:00 PM Afternoon session Panel 3: National level efforts at assessing the prevalence of

violence against women

On what forms of violence against women are data collected at the national level? Where and how is data collected? Is it systematically done? What have these efforts indicated about the prevalence of violence against women in certain countries and what gaps remain? What is the best combination of methods for systematic collection of comparable national level data? How can the experience of international level efforts be integrated into systematic, national level collection of statistics? Holly Johnson (Canada) Elizabeth Ardayfio-Shandorf (Ghana) Dominique Fougeyrollas-Schwebel (France) Ruth Ojiambo Ochieng (Uganda)

(Presentations will be followed by brief questions for clarification)

Presentations on: Work of the Special Rapporteur on Violence Against Women

(Lucinda O�Hanlon) Secretary General�s Study on Violence Against Children (Amaya

Gillespie)

(Coffee break) Synthesis Discussants will synthesize the strengths, weaknesses, limitations and lessons learned from the presentations, as well as identify gaps in terms of methodologies, types of violence against women covered, populations covered, etc. Mary Ellsberg Sylvia Walby

Discussion Tuesday 12 April 2005 10:00 � 11:00 AM Discussion continued Plenary Session 2 Objective: To clarify the data and information needs of users, policy makers and practitioners and identify remaining gaps in the collection of data. 11:00 � 12:30 PM Panel 4: Forms of violence against women for which data is

not systematically collected

Presentations by experts addressing forms of violence against women for which systematic data collection does not exist and is critically needed. How have these forms of violence against women (for example, dowry deaths, acid attacks, early, forced and levirate marriage, sexual violence in times of armed conflict, honour crimes, female genital mutilation/cutting and other forms of violence against women) been documented to date? What are the weaknesses of this type of collection, and how could data collection on these types of violence be systematized, or included in other methodologies of data collection? Sharmeen Farouk (Bangladesh � dowry deaths, acid attacks,

honour crimes) Dalia Farouki (Jordan �honour crimes, other undocumented crimes) Jeanne Ward (Kenya � violence in armed conflict)

(Presentations will be followed by brief questions for clarification)

12:30 � 2:00 PM Lunch

2:00 � 3:15 PM Panel 5: Needs of users of data on violence against women

Presentations by advocates/ practitioners/ service providers highlighting the needs of users of data on violence against women and perspectives from different regions. Asmita Basu (India) Ivy Josiah (Malaysia) Ana Flávia d�Oliveira (Brazil)

(Presentations will be followed by brief questions for clarification)

3:15 � 6:00 PM Discussion (Coffee break) Introduction and formation of working groups Explanation of objectives and expected outputs Wednesday 13 April 2005 10:00 � 6:00 PM Working Groups

a. Use of surveys to assess the prevalence of violence against women: what has worked with surveys that have been used to date and what are the problems? Suggestions for improving survey methodology.

Expected output: Agreement on definitions and types of violence against women that can be included in surveys; development of a short violence against women module and identifying the minimum conditions under which this module could be implemented; elaboration of recommendations for national statistical offices to include questions regarding violence against women in specialized or on-going surveys.

b. Assessing the prevalence of forms of violence for which data and statistics are not

available: what methodologies can be used to document and estimate incidence when it is difficult to identify, reach and/ or survey victims of such violence.

Expected output: Recommendations on how to collect data, document and assess or estimate the prevalence on certain manifestations of violence against women when victims are not easily identified. How can effective policies and strategies for advocacy and intervention be developed to address these forms of violence?

c. Analysis of available crime, health, shelter, death and other demographic data to

assess the prevalence of violence against women: What needs to be considered in order for meaningful data on violence against women to be extrapolated? How should this data ideally be collected and analyzed? How should it be disaggregated?

Expected output: Recommendations on how to improve routine data collection on violence against women and on how to include such data in administrative records

d. Development of indicators on violence against women. What indicators can be used for monitoring at the global level? What kinds of indicators are already being used and how relevant and complete are they in helping to assess trends in violence against women? Should global indicators be developed? What factors need to be addressed when developing global indicators?

Expected output: Agreement on 3-5 indicators for use in global monitoring of violence against women.

Thursday 14 April 2005 Plenary Session 3 10:00 � 12:30 PM Continuation of working groups if needed

Presentations of recommendations and conclusions from working groups.

Questions and discussion Presentation of draft report 12:30 � 2:00 PM Lunch 2:00 � 5:00 PM Finalization and adoption of the report and recommendations for assessing the extent of the problem of violence against women, and on developing methodologies for collecting accurate and thorough data on the prevalence of violence against women globally.